Runoff

SECTION 3

THE THEORY OF THE HYDROLOGY LAYER OF THE EPA SWMM TIER-2 SAW MILL RUN MODELS

 

3.0 Introduction

 

This section describes how the Runoff or Hydrology layer of SWMM transforms the precipitation data from the continuous radar calibrated rainfall data into combined and sanitary inflow.  This inflow enters the trunk sewer and interceptor collection network of SMR and this routing is simulated using the Extran or Hydraulics layer of SWMM.  The Hydraulics layer is described in Section 4.  The actual Runoff layer sanitary sewershed calibration process and the results compared to monitored data collected in the year 2000 are discussed in Section 5.  The calibration process for the combined sewersheds and the results compared to monitored data collected in the year 2001 are discussed in Section 6. 

 

 The important physical processes simulated in the Runoff layer to calibrate the sanitary sewersheds and combined sewersheds are described in this section.  The goal of the hydrology calibration was to use GIS data such as soil maps, sewershed delineation, USGS Digital Elevation Model (DEM) data, land use maps, census data and impervious area coverage as much as possible to minimize parameter calibration and adjustments.  Three key parameters were used to calibrate the combined sewersheds: (1) the percent of the impervious area that is directly connected or the DCIA, (2) a relationship to the overland flow path called the sewershed width that is used for peak flow calibration, and (3) the amount of precipitation that never contributes runoff from the sewershed or the amount of depression storage.  These three parameters are discussed in detail as they work in the Runoff layer and especially as they apply to the Saw Mill Run basin which has unique topographical, structural and hydraulic features.

 

These parameters are considered critical because they address the central issues involved in generating surface runoff from precipitation: (1) how much of the precipitation is available for runoff (the DCIA), (2)  how fast does it runoff (the sewershed width parameter) and (3) how much of the runoff is intercepted before it reaches the inlets of the collection system (depression storage).  

 

The Tier-2 SMR models use the GIS information for the SMR basin to make these parameter estimates easier and supply better data to the model.  Differences between the theory of the Tier-1 models (ALCOSAN, 1996) and the theory of the Tier-2 models (Huber and Dickinson, 1988) are also discussed in this section. 

 

3.1 SWMM Runoff Layer Theory and Data Requirements

 

3.1.1 Components of SWMM

 

The main components of the EPA Stormwater Management Model (SWMM) used in the Tier-2 models of Saw Mill Run (SMR) are the Runoff layer and Extran layers.   The minor components of SWMM that are used in the SMR models (Figure 3.1) are the Transport and Utilities layers (Figure 3.2).  The SMR models use the Temperature, Rain, Combine and Stats Utilities of SWMM to prepare input data for the long term simulation models used for SMR facility planning and prepare the summary tables for the Discharge Monitoring Reports (DMR) quarterly reporting.

                                   

The Runoff layer generates the surface runoff from impervious and pervious surfaces, RDII rainfall-dependent infiltration and inflow (RDII) to the sanitary sewershed pipes and manholes and some of the combined sewershed pipes and manholes, computes snow melt and snow fall, soil infiltration losses from the pervious area of the combined sewersheds and computes the evaporation losses from small surface ponds on the streets, rooftops and open areas (Figure 3.3).

The driving force of the Runoff layer is the precipitation input, which may be a continuous record, a design storm, or a measured storm event.  The storm may be simulated as moving through the watershed or stationary over the watershed.  In the SMR models continuous precipitation data from calibrated radar data is used for each sewershed which lowers the overall model uncertainty by supplying better precipitation estimates per sewershed.

 

The Extran layer will use the flow generated by either the Runoff layer or its own independent user inflow time series to route the flow hydraulically through the storm, sanitary, combined or open channel system.  Extran handles many different boundary conditions at the outfalls, detention tanks, outfall structures, pumps, open and closed conduits, culverts, regulators, bridges, and other specialized types of flow conveyance systems.  In the SMR models Extran integrates the surface runoff from the combined sewersheds, the RDII flow from the sanitary sewersheds and the monitored and population derived dry weather flow to estimate the interceptor flow.  Extran based on the manhole and outfall elevations will estimate the amount of overflow in the SMR network.

 Besides the SWMM model itself, several other software packages are used to produce the model inputs and process model outputs (Figure 3.4).  The model-support components in the Saw Mill Run Runoff/Extran models consist of Arc-View for the generation of the sewershed areas and the delineation of the combined sewersheds into differently sloped areas based on USGS DEM data.  Arc-View will also use soil maps and supply manhole, pipe, and regulator locations.  The CDM RDII Utilities is a standalone program that will allow the estimation of the r, t and k values for a sanitary and combined sewershed based on measured flow data.  The rtk values are then used in the Runoff layer of SWMM to estimate the amount of pipe infiltration or manhole inflow in separate sanitary sewered areas and some combined sewersheds from the rainfall time series.  The three unit hydrographs used in the RDII analysis generate pipe infiltration estimates for the sewersheds from the calibrated radar-rainfall data.  The SWMM model does the actual simulation analysis using the Runoff layer for hydrology and the Extran layer for the pipe, regulator and outfall hydraulics.  The SAS analysis system then takes the predicted and measured flows or heads in the SMR network and performs a statistical comparison of measured storm volumes and predicted storm volumes.

 

 

3.1.2 Hydrologic Model Theory and Data Requirements

 

The SWMM Runoff layer was used to simulate wet-weather flows entering the Saw Mill Run sanitary sewer system from both combined and separate sanitary sewersheds.  RDII infiltration/inflow is estimated in all of the sanitary sewersheds and some of the combined sewersheds.  The combined sewersheds also simulate surface runoff into the inlets and other collection points of the combined drainage network such as rooftop drains.  The Extran layer of SWMM will also read an external time series of dry weather flow as an input to the model.  The flow in the hydraulics network is a combination of these three sources as shown in Table 3.1.

 

Table 3.1 Types of Flow Sources in the SMR Runoff Model

Type of Sanitary System	Base Flow	RDII Flow	Surface Runoff
Combined Sewered Areas	Yes	Yes	Yes
Separate Sewered Areas	Yes	Yes	No

 

For the wet weather calibration phase of the sanitary sewersheds model development in the year 2000 only, the baseflow in the model at each node was taken from the Chester Engineers report of Saw Mill Run (1994).   This baseflow applies only to the year 2000 models which were used for the sanitary calibration.  The SWMM models used after the year 2000 generate the base flow based on monitored data and GIS census data to apportion the monitored base flow among the upstream sewersheds.  The RDII parameters r, t, and k were calibrated using the CDM RDII Utilities program for the monitored sites CS-14, CS-35, CS-42, S-18 and MH.89 during the year 2000.  The rtk parameters computed for these sites were used directly in the model but the rtk parameters for the unmonitored sites were based on the monitored estimates as well as anecdotal evidence about the “leakiness” or “normality” of the trunk sewers.   The unmonitored sanitary sewersheds were classified as “normal” or “leaky” in Section 2 (see Table 2.4).

The Runoff layer parameters were calibrated based on the measured flows at monitored combined sewershed sites S-32, S-40 and S-46 in the SMR basin.  Important calibration parameters were found to be:

n the directly connected impervious area (DCIA),

n impervious area depression storage or interception loss, and

n the sewershed overland flow path length (the subcatchment width parameter in Runoff).

Manning’s roughness values and Green-Ampt infiltration parameters were estimated from literature values and soil maps in Arc-View.   All of the parameters used in modeling surface runoff in the Runoff layer of SWMM are shown in Table 3.2.

 

Table 3.2 Runoff Parameters used in the Runoff Layer of SWMM

Watershed Area Related	Width of subcatchment	Feet
Watershed Area Related	Area of subcatchment	Acres
Watershed Characteristic	Percent imperviousness  (DCIA)	Percent
Watershed Characteristic	Ground slope	(dimensionless)
Watershed Characteristic	Impervious area Manning's roughness.
Watershed Characteristic	Pervious area Manning's roughness.
Watershed Characteristic	Impervious area depression storage	Inches
Watershed Characteristic	Pervious area depression storage	Inches
Infiltration	Average capillary suction of water.	inches
Infiltration	Saturated hydraulic conductivity of soil	Inches/hour.
Infiltration	Initial moisture deficit for soil  volume air/volume voids	(fraction)

 

Goals of the Calibration Process

 

A goal of the calibration process was to utilize as much as possible the information from the GIS.  Sewershed areas, infiltration parameters based on the soil maps of the basin, slope delineation from USGS DEM data, an initial estimate of the impervious combined sewershed coverage and an initial estimate of the sewershed width parameter were all generated from Arc-View. The calibration parameters for the combined sewersheds were restricted to an adjustment of the sewershed width, the impervious percent of the sewershed that was directly connected to the combined drainage system (DCIA) and the combined sewershed depression storage.  All of these three parameters were calibrated based on monitored flows at sites S-32, S-40 and S-46 in the SMR basin during the year 2001.  

Another goal of the calibration process was to use reasonable assumptions about the model parameters and not try to use different values of the same parameter for various monitored sites unless dictated by the monitoring data   For example, the same value of the percent of the impervious area that immediate runoff into the inlets was used for all of the combined sewersheds because having variable values per sewershed was not warranted by the data or justifiable from site inspections.

 

Saw Mill Run Water Balance for the Year 2000 and Year 2001 Calibrations

 

The following sections discuss the components of the Saw Mill Run water balance for the period between March and October 2000 as an example of the major pathways in the SMR hydrologic cycle (Figure 3.6) and the combined sewershed calibration period between April and July, 2001 as another example (Figure 3.8).   The water balance of a watershed is a quantitative description of the components of the hydrologic cycle (Figure 3.7).  For a given system, the hydrologic inputs and outputs must be equal.  Defining the system of interest as the Saw Mill Run surface water system, the input to the system is precipitation and the outputs are soil infiltration, evaporation, and runoff (including snowmelt) reaching the sewershed.  Figure 3.7 shows these components during the calibration period in the year 2000 as represented in the hydrologic model.   In addition to the typical input and outputs shown in Figure 3.8 there was 2.3 inches of evaporation from the combined sewersheds of SMR during the